Industrial-Scale Machine Learning
summary
AWS claims customers run over a million labeling tasks per day and serve 100 billion+ predictions monthly while some deploy millions of personalized models, so SageMaker focuses on industrial workflows: SageMaker Multi-Model Endpoints let one endpoint host many models with sharding, pin-in-memory and autoscaling; a model registry plus feature store records versions and reproducible data snapshots; Model Monitor captures requests to S3, merges ground truth by ID and computes alerts (confusion, RMSE, custom metrics); SageMaker Clarify runs kernel SHAP on a shadow endpoint for explainability at scale.
AI-written summary · about
chapters (8)
- 0:00Industrial ML governance
- 2:36ML as present necessity
- 5:23Industrialization of machine learning
- 7:28Model deployment & monitoring
- 10:59SageMaker multimodal endpoints
- 13:10Monitoring models at scale
- 16:02Bias & explainability with SageMaker Clarify
- 18:12Automating retraining & alerts
related talks
33mBuild, train, and deploy ML models using Amazon SageMaker
32mBuild, Train, and Deploy Machine Learning Models on AWS
37mStreamlining Machine Learning Operations
27mScaling ML to the next level: Hosting thousands of models on SageMaker
34mLat Designing and Building Complex Machine Learning....
37mStreamlining Machine Learning Operations
34mBuild, train and deploy Machine Learning models with Amazon SageMaker
19mUsing Amazon SageMaker to develop and deploy ML models